Presidential election, 2016

[pbrower2a]

Map showing likelihood of wins for Clinton and Trump, Johnson considered

Note: this is likelihood and not margin. Margin may be related to likelihood at this stage.



win
chance        sat

99%+       9
95-98.9    8
90-94.9    7
80-89.9    6
70-79.9    5
60-69.9    4
55-59.9    3
52-54.9    2


At this stage I would predict Obama 2012 + Arizona, North Carolina, and the Second Congressional District of Nebraska.

It is really bad for Donald Trump. The only electoral vote that he has a now-meaningful chance to pick up that has gone consistently to Democratic nominees for President since 1992 is the Second Congressional District of Maine.

Data from this source.

Going by chances involving states strictly within the Blue (on my source the traditional Red) Firewall I get

state  win-chance cumulative electoral votes
DC    >99.9%          3
MD      99.8             13
HI        99.6            17
MA       98.8            27
NY       98.7            56

CA       98.6           111
IL         96.9           131
VT       96.7            134
RI        95.3           138
DE       93.2           142
WA      91.9           154
NJ        91.6           168  
MI        90.7           184 
CT       89.2           192
NM       87.1           197

WI       86.0           207
MN       85.8           217
OR       85.2           224
ME       82.9           227 (not including the 2nd congressional district, which is vulnerable)
PA        81.3           247

Trump could win comfortably by losing only states that haven't gone for a Republican nominee since 1988 and New Mexico, which fits better in this list than anything else.

I am going out of sequence here. Unless Donald Trump wins the state on this list that he has the greatest chance of winning or can pick off Oregon, Minnesota, Wisconsin, or Connecticut (you can forget any state 'higher' on the list) or Maine and New Mexico together, there is one state that he absolutely must win:

FLORIDA -- which Hillary Clinton has a 73.7% chance of winning.

Ouch.

As for the states above -- except for New Mexico these states are concentrated along the east coast to the north and east of the Potomac, the Great Lakes, and the West Coast (OK, Hawaii is far off California) and generally have some similarities of culture and economics.

Now let's look at some other states:

VA    78.6  260
NV    75.9  266
NH    74.6  270

Virginia, Nevada, and New Hampshire put Hillary Clinton at the magic 270. Replace Nevada with Iowa, and Hillary Clinton still wins. So if Hillary Clinton wins every state that she has at least a 74% chance of winning, she wins the Presidency.

FL     73.7  299 (already mentioned)
OH    70.8  317 (tipping-point state in 2004)

IA      70.1 323 (tipping-point state in 2008)
CO    68.3  332 (tipping-point state in 2012)

Figure that ME-02 is in there. This is identical to Obama 2012.

...This is a simple model. It says nothing about the character of the states. A Republican is going to lose Iowa if he says  stupid stuff about agriculture. Barack Obama knew enough about agriculture to win handily in Illinois, a very rural state outside of Greater Chicago. I saw some of his literature directed at farmers, and much of it was an excoriation of meth. Meth-makers were stealing anhydrous ammonia (a favorite fertilizer) and often letting the rest gravitate to whatever happened to be below. If whatever happened to be below was livestock, then the ammonia killed the livestock. Obama promised a crackdown on meth, and we have it the crackdown on meth. American farmers who don't like their anhydrous ammonia stolen or let loose to kill livestock can thank him. It may have swung some votes in Ohio.

I question whether Donald Trump can say anything wise about agriculture.

Trump absolutely needs these states, but he is at greater risk of losing them than of winning them:

NC    60.5   347
AZ    55.0   359
Nebraska's Second Congressional District (most of Greater Omaha) would put Hillary Clinton at 360.

At  360 electoral votes, a Presidential winner is on the brink of a landslide. Obama won a regional landslide in 2008.

It's not until Missouri that Trump finally finds a state in which he now has a better chance of winning (52%) than losing. But sticking with percentages for Hillary Clinton,

MO    47.7   369
GA     43.3   385
SC     39.1   393


The drop-off is sharp from there.

MS   32.5
KS   32.3
SD   29.8
IN    29.5
TX    28.5
ND   28.2
AS    26.8
MT    26.3
UT    24.4

I may have missed something here -- but at this stage, no Republican should have practically one chance in three of losing Kansas or one chance in four of losing Texas or Utah in a statewide election. Donald Trump has a bigger chance of losing Texas than he does of losing Florida, and he has a bigger chance of losing Kansas than of winning Ohio.

Silver's model cannot predict a collapse of a campaign. It can only recognize it as it happens.